Question

我有三个数据框。我想从所有3个数据框中按日期2020-10-26选择行，然后我要创建一个新的数据框。怎么做？

import pandas as pd
from datetime import date

date1 = date.fromisoformat('2020-10-25')
date2 = date.fromisoformat('2020-10-26')
date3 = date.fromisoformat('2020-10-27')

for x in range(3):
    df = pd.DataFrame([(date1,f'Stu{x}j', x+1, f'Varan{x}j', x*400+2), 
                        (date2,f'aja{x}k', x+2, f'Del{x}j', x*634+3), 
                        (date3,f'Aadi{x}t', x+4, f'Mumb{x}j', x*454+4),
                        (date2,f'har{x}h', x+5, f'bom{x}j', x*124+5)],
                       columns =['Date','Name', 'Age',  
                             'City', 'Salary'])
    df.set_index('Date', inplace = True)
    print(df)

Answer 1

以下代码将为您提供帮助

import pandas as pd
from datetime import date

date1 = date.fromisoformat('2020-10-25')
date2 = date.fromisoformat('2020-10-26')
date3 = date.fromisoformat('2020-10-27')
res = pd.DataFrame()
for x in range(3):
    df = pd.DataFrame([(date1,f'Stu{x}j', x+1, f'Varan{x}j', x*400+2), 
                        (date2,f'aja{x}k', x+2, f'Del{x}j', x*634+3), 
                        (date3,f'Aadi{x}t', x+4, f'Mumb{x}j', x*454+4),
                        (date2,f'har{x}h', x+5, f'bom{x}j', x*124+5)],
                       columns =['Date','Name', 'Age',  
                             'City', 'Salary'])
    temp = df[df['Date'].isin([date2])]
    res = res.append(temp)
print(res)

旧方法：

fr = [df1, df2, df3]
full = pd.concat(fr).reset_index()
full[full.Date.isin([date2])]

Answer 2

您可以这样做：

row1 = df1.loc[date2, :]
row2 = df2.loc[date2, :]
row3 = df3.loc[date2, :]
pd.concat([row1,row2,row3], axis=1).T

输出：

            Name    Age City        Salary
2020-10-26  ajay    32  Delhi       25000
2020-10-26  Saumya  32  Kanpur      20000
2020-10-26  Aaditya 40  Dehradun    24000

从多数据框熊猫的选定行中创建新的数据框

2 个答案: